CN111814840A - Method, system, equipment and medium for evaluating quality of face image - Google Patents

Abstract

The present invention provides a face image quality assessment method, system, equipment and medium, including: when the total face score corresponding to the face image is within a preset threshold range, the scores obtained according to the obtained face quality assessment models , and output the face image quality assessment result; the present invention greatly improves the speed and efficiency of the face image quality assessment by adjusting the quality assessment strategy.

Description

A face image quality assessment method, system, device and medium

technical field

The present invention relates to the fields of artificial intelligence and image processing, and in particular, to a method, system, device and medium for evaluating the quality of a face image.

Background technique

With the development of artificial intelligence, face recognition technology is more and more widely used in real life. Face quality assessment is a very important part of the face recognition process. How to quickly and accurately give the face quality score is a difficult problem.

The existing face quality assessment needs to calculate the results of the key points, sharpness, brightness, angle, total score, etc., and take the weighted average of these scores to obtain the total score. The evaluation time is positively related to the number of faces contained in the picture. . Therefore, in the case of a large flow of people, such as 100 people in a certain period of time in the video stream, then the snapshots of multiple consecutive frames will contain 100 faces, and the face quality evaluation will take a long time, which will affect the face recognition. efficiency.

SUMMARY OF THE INVENTION

In view of the above problems in the prior art, the present invention proposes a face image quality assessment method, system, device and medium, which mainly solves the problem that the processing speed of the traditional quality assessment method is slow in processing the data of large traffic.

In order to achieve the above objects and other objects, the technical solutions adopted in the present invention are as follows.

A face image quality assessment method, comprising:

When the total face score corresponding to the face image is within the preset threshold range, the face image quality assessment result is output according to the scores obtained by each face quality assessment model.

Optionally, when the face total score corresponding to the face image exceeds the preset threshold range, output the face image quality assessment result according to the face total score.

Optionally, the total face score corresponding to the face image is obtained in advance according to a quality evaluation model for the total face score.

Optionally, the face total score quality assessment model identifies the pre-marked face image feature regions, and outputs the face total score.

Optionally, for the multi-frame images of the same target object, according to the determined frame interval number, determine the target frame to be evaluated;

Through the face quality assessment model, quality assessment is performed on the face image of the target object contained in the target frame to be assessed.

Optionally, the process of determining the number of frame intervals is:

Set the initial frame interval number, and obtain the interval frame of the same target object according to the initial frame interval number; Obtain the face image quality error of the same target object in the interval frame; Determine whether the quality error is within the allowable setting of the setting within the range;

If it exceeds the allowable range, adjust the initial frame interval number until the quality score error is within the allowable range.

Optionally, the face quality assessment model includes at least:

Face total score quality assessment model, face key point quality assessment model, face definition quality assessment model, face angle quality assessment model, face brightness quality assessment model.

Optionally, the threshold range is determined according to a preset range of image quality parameters:

If the total score of the human face is less than the lower bound of the preset range, then the total score of the human face is used as the lower bound of the threshold range;

If the total score of the human face is greater than the upper bound of the preset range, the total score of the human face is used as the upper bound of the threshold range.

Optionally, the image quality parameter includes at least one of the following: an image brightness value, an image gray value, and an image sharpness value.

Optionally, when the total face score is greater than the upper bound of the threshold range, obtain the error between the total face score plus a preset constant and the comprehensive score of all face quality assessment models, and judge the error. Whether it is less than the preset constant, and if so, adding the preset constant to the total face score is used to update the upper bound of the threshold range.

Optionally, weighting is performed on the scores of each face quality assessment model to obtain the comprehensive score.

A face image quality assessment system, comprising:

The quality assessment module is used to output the quality assessment result of the face image according to the scores obtained by the obtained face quality assessment models when the total face score corresponding to the face image is within the preset threshold range.

Optionally, when the face total score corresponding to the face image exceeds the preset threshold range, output the face image quality assessment result according to the face total score.

Optionally, a total score evaluation module is included, which is used to obtain the total face score corresponding to the face image according to the total score quality evaluation model of the human face in advance.

Optionally, the face total score quality assessment model identifies the pre-marked face image feature regions, and outputs the face total score.

Optionally, include interval evaluation module, for the multi-frame image of same target object, according to the frame interval number determined, determine the target frame to be evaluated;

Through the face quality assessment model, quality assessment is performed on the face image of the target object contained in the target frame to be assessed.

Optionally, the process of determining the number of frame intervals is:

Set the initial frame interval number, and obtain the interval frame of the same target object according to the initial frame interval number; Obtain the face image quality error of the same target object in the interval frame; Determine whether the quality error is within the allowable setting of the setting within the range;

If it exceeds the allowable range, adjust the initial frame interval number until the quality score error is within the allowable range.

Optionally, the face quality assessment model includes at least:

Face total score quality assessment model, face key point quality assessment model, face definition quality assessment model, face angle quality assessment model, face brightness quality assessment model.

Optionally, the process of determining the preset threshold range is:

The threshold range is determined according to the preset range of image quality parameters:

If the total score of the human face is less than the lower bound of the preset range, then the total score of the human face is used as the lower bound of the threshold range;

If the total score of the human face is greater than the upper bound of the preset range, the total score of the human face is used as the upper bound of the threshold range.

Optionally, the image quality parameter includes at least one of the following: an image brightness value, an image gray value, and an image sharpness value.

Optionally, a threshold update module is included, for when the total score of the human face is greater than the upper bound of the threshold range, the total score of the human face is obtained after adding a preset constant and the comprehensive score of all face quality evaluation models. The error is determined, whether the error is smaller than the preset constant, and if so, the total face score plus the preset constant is used to update the upper bound of the threshold range.

Optionally, a comprehensive scoring module is included to perform weighting processing on the scores of each face quality assessment model to obtain the comprehensive score.

A device comprising:

one or more processors; and

One or more machine-readable media having instructions stored thereon, when executed by the one or more processors, cause the apparatus to perform the method for assessing the quality of a human face image.

One or more machine-readable media having instructions stored thereon that, when executed by one or more processors, cause an apparatus to perform the described method for assessing the quality of a human face image.

As described above, the present invention provides a method, system, medium and device for evaluating the quality of a face image, which have the following beneficial effects.

Evaluate face image quality through total face score, reduce unnecessary model evaluation calculations, reduce CPU usage, and improve quality evaluation efficiency.

Description of drawings

FIG. 1 is a flowchart of a method for evaluating the quality of a face image according to an embodiment of the present invention.

FIG. 2 is a block diagram of a face image quality assessment system according to an embodiment of the present invention.

FIG. 3 is a schematic structural diagram of a terminal device in an embodiment of the present invention.

FIG. 4 is a schematic structural diagram of a terminal device in another embodiment of the present invention.

Detailed ways

The embodiments of the present invention are described below through specific specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other under the condition of no conflict.

It should be noted that the drawings provided in the following embodiments are only used to illustrate the basic concept of the present invention in a schematic way, so the drawings only show the components related to the present invention rather than the number, shape and number of components in actual implementation. For dimension drawing, the type, quantity and proportion of each component can be arbitrarily changed in actual implementation, and the component layout may also be more complicated.

Referring to FIG. 1, the present invention provides a method for evaluating the quality of a face image, which includes, when the total face score corresponding to the face image is within a preset threshold range, according to the scores obtained by the obtained face quality evaluation models, outputting Face image quality assessment results.

In one embodiment, a variety of face quality assessment models can be pre-trained, wherein the face quality assessment models include at least a face total score quality assessment model, a face key point quality assessment model, a face clarity quality assessment model, and a human face quality assessment model. Face angle quality assessment model, face brightness quality assessment model. According to each face quality assessment model, the quality of face images is scored separately.

Specifically, when training the face total score scoring model, the training sample images can be marked with faces in advance, and the marked face image feature areas can be identified by using a face recognition algorithm. For example, the face image features can be identified. The area is compared with the standard face template, and the sharpness of key feature points in the feature area, the ambiguity and brightness of the feature area, etc. are judged, and the total score of the face is output. A face-to-face total score evaluation model trained on a large number of sample images. Among them, the face recognition algorithm can use deep neural network algorithm, feature-based face recognition algorithm, template-based face recognition algorithm, etc.

When scoring the quality of the key feature points of the face, a key point scoring model can be established through a deep convolutional neural network, and the facial features such as eyes, eyebrows, nose, mouth and other facial features can be detected and scored respectively, and the scores of each key feature point. Weighted average to obtain the scoring results of the keypoint scoring model.

During sharpness scoring, the sharpness scoring model can be trained by conventional sharpness calculation methods such as grayscale variance algorithm (SMD), energy gradient algorithm, and Laplace algorithm. The Laplace operator is convolved with the pixels in the face image, and the gradient values in the horizontal and vertical directions of the face image are extracted respectively. When a bright spot appears in the darker area of the face image, the Laplacian operator is used. The connected points can be made brighter, and the sharpness is scored by the calculated number of bright spots and the size of the bright spot area, and the scoring result of the sharpness scoring model is obtained. Among them, the Laplacian can be expressed as follows.

0 1 0 1 -4 1 0 1 0

When performing angle scoring, a training sample set can be constructed. The sample set contains the frontal image and side image of the corresponding face, respectively. Through conventional neural network algorithms such as BP neural network, an error function is constructed based on the deviation of the angle of the frontal image and the side image. Perform model training to obtain an angle score model. The smaller the error function value, the higher the angle score.

When scoring brightness, the frequency of occurrence of all pixels in the face image can be counted according to the gray scale by the gray histogram method. The gray scale range can be set to 80-200, and the pixels within this gray level range can be counted The frequency of occurrence, and the brightness of the face image is scored according to the statistical frequency. The higher the frequency, the higher the brightness score.

After each face quality assessment model is obtained, the face in the video stream can be scored only through the face total score quality assessment model in advance, and the total face score can be obtained.

In one embodiment, the threshold range of the total face score can be set, and specifically, the upper/lower bounds of the threshold range of the total face score can be evaluated through the image quality parameter. The image quality parameters may include image brightness value, image gray value, image sharpness value, and the like. Image quality parameters can be obtained by using corresponding identification algorithms such as gray mean normalization method. The preset range of image quality parameters can be obtained in advance. When the total face score is less than the lower bound of the preset range, it is determined that the quality of the face image is poor, and the lower bound of the preset range is used as the lower bound of the threshold range of the total face score; When the total face score is greater than the upper bound of the preset range, it is determined that the quality of the face image is good, and the upper bound of the preset range is taken as the upper bound of the threshold range of the total score of the face. In another embodiment, the comprehensive scores of the sample images can also be counted through each face quality assessment model in advance. Specifically, the weight of each face quality assessment model can be set, and the comprehensive score of each face quality assessment model can be obtained by using a weighted average. score. And set the value range of the comprehensive score to distinguish images of good quality and poor quality. When the total score of the sample image is less than the lower bound of the comprehensive score, the image quality is poor, and the lower bound of the comprehensive score is directly used as the total score of the face. The lower bound of the threshold range; when the total face score of the sample image is greater than the upper bound of the comprehensive score, the image quality is good, and the upper bound of the comprehensive score is directly used as the upper bound of the threshold range of the total score of the face.

The specific threshold range value and determination method can be flexibly selected according to the actual application scenario.

In one embodiment, in practical applications, when the total face score is greater than the upper bound of the threshold range, a preset constant can be added to the total face score to calculate the comprehensive score of each face quality assessment model and add The error between the total score of the face after the preset constant is set. If the confidence level of the error is less than the preset constant is 95%, the sum of the total score of the face and the preset constant is used as the upper bound of the threshold range, and the human face is updated. The upper bound of the threshold range for the total face score.

In one embodiment, when the total face score of the face image to be evaluated exceeds the set threshold range, the total face score can be used as the face quality evaluation result, and the face images below the lower bound of the threshold range are filtered out. Or select high-quality face images above the upper bound of the threshold range.

When the total face score of the face image to be evaluated is within the threshold range, the result obtained by the weighted average of the scores of each face quality evaluation model is used as the quality evaluation result of the face image to be evaluated.

In one embodiment, when the face quality assessment is performed on the captured video stream, the face of the same person in consecutive multiple frames of images can be tracked according to the face ID. According to the determined number of frame intervals, multiple target frame images to be evaluated are determined from multiple frames of faces of the same person. The face quality assessment is only performed on the multi-frame face images obtained according to the determined frame interval, so as to improve the face quality assessment speed.

In one embodiment, the determination of the frame interval may include the following steps:

Set the initial frame interval number, and obtain the interval frame of the same target object according to the initial frame interval number; Obtain the face image quality error of the same target object in the interval frame; Determine whether the quality error is within the allowable setting of the setting within the range;

If it exceeds the allowable range, adjust the initial frame interval number until the quality score error is within the allowable range. Specifically, it can be set that the continuous frame images of the same person are 7 frames, and the initial frame interval is 2, that is, the face frame images of the same person are obtained every two frames and the face quality assessment is performed according to the aforementioned steps. The calculated error between the quality scores of the first frame, the fourth frame and the seventh frame image, if the allowable error range is set to [-0.1, 0.1], then when the statistical confidence is 95%, the Whether the error of the quality score falls within the allowable range, if it is within the allowable range, the interval frame images of the same person can be obtained according to the initial frame interval for quality scoring; if it exceeds the allowable range, reduce the frame interval and recalculate the error, until the error falls within the allowable range.

Referring to Fig. 2, this embodiment provides a system for evaluating the quality of a human face image, which is used to perform the method for evaluating the quality of a human face image described in the foregoing method embodiments. Since the technical principles of the system embodiments are similar to the technical principles of the foregoing method embodiments, the same technical details will not be repeated.

In one embodiment, a face image quality assessment system includes: a quality assessment module configured to assess the quality of each face according to the obtained face quality assessment model when the total face score corresponding to the face image is within a preset threshold range. The obtained score, output the face image quality evaluation result.

Optionally, when the face total score corresponding to the face image exceeds the preset threshold range, output the face image quality assessment result according to the face total score.

Optionally, a total score evaluation module is included, which is used to obtain the total face score corresponding to the face image according to the total score quality evaluation model of the human face in advance.

Optionally, the face total score quality assessment model identifies the pre-marked face image feature regions, and outputs the face total score.

Optionally, include interval evaluation module, for the multi-frame image of same target object, according to the frame interval number determined, determine the target frame to be evaluated;

Through the face quality assessment model, quality assessment is performed on the face image of the target object contained in the target frame to be assessed.

Optionally, the process of determining the number of frame intervals is:

Set the initial frame interval number, and obtain the interval frame of the same target object according to the initial frame interval number; Obtain the face image quality error of the same target object in the interval frame; Determine whether the quality error is within the allowable setting of the setting within the range;

If it exceeds the allowable range, adjust the initial frame interval number until the quality score error is within the allowable range.

Optionally, the face quality assessment model includes at least:

Face total score quality assessment model, face key point quality assessment model, face definition quality assessment model, face angle quality assessment model, face brightness quality assessment model.

Optionally, the process of determining the preset threshold range is:

The threshold range is determined according to the preset range of image quality parameters:

If the total score of the human face is less than the lower bound of the preset range, then the total score of the human face is used as the lower bound of the threshold range;

If the total score of the human face is greater than the upper bound of the preset range, the total score of the human face is used as the upper bound of the threshold range.

Optionally, the image quality parameter includes at least one of the following: an image brightness value, an image gray value, and an image sharpness value.

Optionally, a threshold update module is included, for when the total score of the human face is greater than the upper bound of the threshold range, the total score of the human face is obtained after adding a preset constant and the comprehensive score of all face quality evaluation models. The error is determined, whether the error is smaller than the preset constant, and if so, the total face score plus the preset constant is used to update the upper bound of the threshold range.

Optionally, a comprehensive scoring module is included to perform weighting processing on the scores of each face quality assessment model to obtain the comprehensive score.

Embodiments of the present application also provide a device, which may include: one or more processors; and one or more machine-readable media on which instructions are stored, when executed by the one or more processors At the time, the device is caused to execute the method described in FIG. 1 . In practical applications, the device can be used as a terminal device or a server. Examples of terminal devices can include: smart phones, tablet computers, e-book readers, MP3 (Motion Picture Experts Compression Standard Voice Layer 3, Moving Picture Experts Group Audio Layer III) players, MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop computers, car computers, desktop computers, set-top boxes, smart TVs, wearable devices Etc., the embodiments of the present application do not limit specific devices.

Embodiments of the present application further provide a non-volatile readable storage medium, where one or more modules (programs) are stored in the storage medium, and when the one or more modules are applied to a device, the device can be executed by the device. Instructions for steps included in the method for evaluating the quality of a face image in FIG. 1 according to the embodiment of the present application.

FIG. 3 is a schematic diagram of a hardware structure of a terminal device according to an embodiment of the present application. As shown, the terminal device may include: an input device 1100, a first processor 1101, an output device 1102, a first memory 1103, and at least one communication bus 1104. A communication bus 1104 is used to enable communication connections between elements. The first memory 1103 may include a high-speed RAM memory, and may also include a non-volatile storage NVM, such as at least one disk memory, and various programs may be stored in the first memory 1103 for completing various processing functions and implementing this embodiment. method steps.

Optionally, the first processor 1101 may be, for example, a central processing unit (Central Processing Unit, CPU for short), an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable A logic device (PLD), field programmable gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation, the processor 1101 is coupled to the aforementioned input device 1100 and output device 1102 through wired or wireless connections .

Optionally, the above-mentioned input device 1100 may include a variety of input devices, for example, may include at least one of a user-oriented user interface, a device-oriented device interface, a software programmable interface, a camera, and a sensor. Optionally, the device-oriented device interface may be a wired interface for data transmission between devices, or a hardware plug-in interface (such as a USB interface, serial port, etc.) for data transmission between devices. ); optionally, the user-oriented user interface may be, for example, a user-oriented control button, a voice input device for receiving voice input, and a touch-sensing device (such as a touch screen with a touch-sensing function, a touch-sensitive device for receiving user touch input) Control panel, etc.); Optionally, the programmable interface of the above-mentioned software can be, for example, an entry for users to edit or modify programs, such as an input pin interface or an input interface of a chip, etc.; the output device 1102 can include output devices such as a display and audio .

In this embodiment, the processor of the terminal device includes functions for executing each module of the speech recognition device in each device, and the specific functions and technical effects may refer to the above-mentioned embodiments, which will not be repeated here.

FIG. 4 is a schematic diagram of a hardware structure of a terminal device according to another embodiment of the present application. Fig. 4 is a specific embodiment of Fig. 3 in the implementation process. As shown in the figure, the terminal device in this embodiment may include a second processor 1201 and a second memory 1202.

The second processor 1201 executes the computer program code stored in the second memory 1202 to implement the method described in FIG. 1 in the above embodiment.

The second memory 1202 is configured to store various types of data to support operations at the terminal device. Examples of such data include instructions for any application or method operating on the end device, such as messages, pictures, videos, etc. The second memory 1202 may include random access memory (random access memory, RAM for short), and may also include non-volatile memory (non-volatile memory), such as at least one disk storage.

Optionally, the first processor 1201 is provided in the processing component 1200 . The terminal device may further include: a communication component 1203, a power supply component 1204, a multimedia component 1205, a voice component 1206, an input/output interface 1207 and/or a sensor component 1208. Components and the like specifically included in the terminal device are set according to actual requirements, which are not limited in this embodiment.

The processing component 1200 generally controls the overall operation of the terminal device. The processing component 1200 may include one or more second processors 1201 to execute instructions to perform all or some of the steps of the method shown in FIG. 1 above. Additionally, processing component 1200 may include one or more modules that facilitate interaction between processing component 1200 and other components. For example, processing component 1200 may include a multimedia module to facilitate interaction between multimedia component 1205 and processing component 1200.

Power component 1204 provides power to various components of the terminal device. Power components 1204 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to end devices.

The multimedia component 1205 includes a display screen that provides an output interface between the terminal device and the user. In some embodiments, the display screen may include a liquid crystal display (LCD) and a touch panel (TP). If the display screen includes a touch panel, the display screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, swipe, and gestures on the touch panel. The touch sensor may not only sense the boundaries of a touch or swipe action, but also detect the duration and pressure associated with the touch or swipe action.

Speech component 1206 is configured to output and/or input speech signals. For example, the speech component 1206 includes a microphone (MIC) that is configured to receive external speech signals when the terminal device is in an operational mode, such as a speech recognition mode. The received speech signal may be further stored in the second memory 1202 or transmitted via the communication component 1203. In some embodiments, the speech component 1206 also includes a speaker for outputting speech signals.

The input/output interface 1207 provides an interface between the processing component 1200 and a peripheral interface module, which may be a click wheel, a button, or the like. These buttons may include, but are not limited to: volume buttons, start button, and lock button.

Sensor assembly 1208 includes one or more sensors for providing various aspects of the status assessment for the end device. For example, the sensor assembly 1208 can detect the open/closed state of the end device, the relative positioning of the assembly, the presence or absence of user contact with the end device. The sensor assembly 1208 may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact, including detecting the distance between the user and the end device. In some embodiments, the sensor assembly 1208 may also include a camera or the like.

Communication component 1203 is configured to facilitate wired or wireless communications between end devices and other devices. Terminal devices can access wireless networks based on communication standards, such as WiFi, 2G or 3G, or a combination thereof. In one embodiment, the terminal device may include a SIM card slot, and the SIM card slot is used for inserting a SIM card, so that the terminal device can log in to the GPRS network and establish communication with the server through the Internet.

As can be seen from the above, the communication component 1203, the voice component 1206, the input/output interface 1207, and the sensor component 1208 involved in the embodiment of Fig. 4 can all be implemented as the input device in the embodiment of Fig. 3 .

To sum up, the present invention provides a method, system, device and medium for evaluating the quality of a face image. When the total face score threshold is exceeded, the total face score is selected as an influencing factor, which simplifies the scoring process and improves the scoring efficiency. ;Adjusting the total face score threshold through each face quality scoring model and estimating the total face score threshold through the error can ensure that the better image has a higher quality score and the obtained quality score is within the acceptable error range. Within; by setting the frame interval to reduce the image data involved in the quality assessment, the time-consuming face quality assessment can be greatly reduced. Therefore, the present invention effectively overcomes various shortcomings in the prior art and has high industrial application value.

The above-mentioned embodiments merely illustrate the principles and effects of the present invention, but are not intended to limit the present invention. Any person skilled in the art can modify or change the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or changes made by those with ordinary knowledge in the technical field without departing from the spirit and technical idea disclosed in the present invention should still be covered by the claims of the present invention.

Claims (24)

1. A method for evaluating the quality of a face image is characterized by comprising the following steps:

and when the total face score corresponding to the face image is within a preset threshold range, outputting a face image quality evaluation result according to the score obtained by each obtained face quality evaluation model.

2. The face image quality evaluation method according to claim 1,

and when the total face score corresponding to the face image exceeds the preset threshold range, outputting a face image quality evaluation result according to the total face score.

3. The method for evaluating the quality of a face image according to claim 1, characterized in that the face total score corresponding to the face image is obtained in advance according to a face total score quality evaluation model.

4. The method for evaluating the quality of the face image according to claim 3, wherein the face total score quality evaluation model identifies a pre-labeled face image characteristic region and outputs a face total score.

5. The face image quality evaluation method according to claim 1,

determining a target frame to be evaluated according to the determined frame interval number for the multi-frame images of the same target object;

and performing quality evaluation on the face image of the target object contained in the target frame to be evaluated through the face quality evaluation model.

6. The method for evaluating the quality of a face image according to claim 5, wherein the process of determining the number of frame intervals is:

setting an initial frame interval number, and acquiring interval frames of the same target object according to the initial frame interval number; acquiring the quality difference error of the face image of the same target object in the interval frame; judging whether the mass component error is within a set allowable range;

and if the quality score exceeds the allowable range, adjusting the initial frame interval number until the quality score error is within the allowable range.

7. The method according to claim 1, wherein the face image quality evaluation model at least comprises:

the system comprises a face total score quality evaluation model, a face key point quality evaluation model, a face definition quality evaluation model, a face angle quality evaluation model and a face brightness quality evaluation model.

8. The face image quality evaluation method according to claim 7,

determining the threshold range according to the preset range of the image quality parameters:

if the total face score is smaller than the lower bound of the preset range, taking the total face score as the lower bound of the threshold range;

and if the total face score is larger than the upper bound of the preset range, taking the total face score as the upper bound of the threshold range.

9. The method according to claim 8, wherein the image quality parameter comprises at least one of: image brightness value, image gray scale value and image definition value.

10. The method according to claim 8, wherein when the total score of the face is greater than the upper bound of the threshold range, an error between the total score of the face plus a preset constant and the comprehensive scores of all face quality evaluation models is obtained, whether the error is smaller than the preset constant is judged, and if yes, the total score of the face plus the preset constant is used for updating the upper bound of the threshold range.

11. The method according to claim 10, wherein the scores of the face quality evaluation models are weighted to obtain the composite score.

12. A face image quality evaluation system, comprising:

and the quality evaluation module is used for outputting the quality evaluation result of the face image according to the obtained scores obtained by the face quality evaluation models when the total face score corresponding to the face image is within the preset threshold range.

13. The facial image quality evaluation system according to claim 12,

and when the total face score corresponding to the face image exceeds the preset threshold range, outputting a face image quality evaluation result according to the total face score.

14. The system for evaluating the quality of a human face image according to claim 12, comprising a total score evaluation module for obtaining a total score of a human face corresponding to the human face image in advance according to a total score quality evaluation model of the human face.

15. The system for evaluating the quality of a facial image according to claim 14, wherein the model for evaluating the quality of the total score of a facial image identifies a characteristic region of a facial image labeled in advance and outputs the total score of the facial image.

16. The facial image quality evaluation system of claim 12, characterized by comprising an interval evaluation module for determining a target frame to be evaluated according to the determined frame interval number for a plurality of frames of images of the same target object;

and performing quality evaluation on the face image of the target object contained in the target frame to be evaluated through the face quality evaluation model.

17. The system for evaluating the quality of a human face image according to claim 16, wherein the process of determining the number of frame intervals is:

setting an initial frame interval number, and acquiring interval frames of the same target object according to the initial frame interval number; acquiring the quality difference error of the face image of the same target object in the interval frame; judging whether the mass component error is within a set allowable range;

and if the quality score exceeds the allowable range, adjusting the initial frame interval number until the quality score error is within the allowable range.

18. The facial image quality evaluation system according to claim 12, wherein said facial quality evaluation model comprises at least:

the system comprises a face total score quality evaluation model, a face key point quality evaluation model, a face definition quality evaluation model, a face angle quality evaluation model and a face brightness quality evaluation model.

19. The system for evaluating the quality of a human face image according to claim 18, wherein the predetermined threshold range is determined by:

determining the threshold range according to the preset range of the image quality parameters:

if the total face score is smaller than the lower bound of the preset range, taking the total face score as the lower bound of the threshold range;

and if the total face score is larger than the upper bound of the preset range, taking the total face score as the upper bound of the threshold range.

20. The facial image quality evaluation system according to claim 19, wherein said image quality parameters comprise at least one of: image brightness value, image gray scale value and image definition value.

21. The system according to claim 19, comprising a threshold updating module, configured to, when the total score of the face is greater than the upper bound of the threshold range, obtain an error between the total score of the face and a combined score of all face quality evaluation models after adding a preset constant, determine whether the error is smaller than the preset constant, and if so, add the preset constant to the total score of the face to update the upper bound of the threshold range.

22. The system according to claim 21, comprising a comprehensive scoring module configured to perform weighting processing on the scores of the respective face quality evaluation models to obtain the comprehensive scores.

23. An apparatus, comprising:

one or more processors; and

one or more machine-readable media having instructions stored thereon that, when executed by the one or more processors, cause the apparatus to perform the method recited by one or more of claims 1-11.

24. One or more machine-readable media having instructions stored thereon, which when executed by one or more processors, cause an apparatus to perform the method recited by one or more of claims 1-11.

Images